All products

Storyboard a reel with AI

Agree the still frames of a vertical video before paying for new motion. Keep dialogue, captions and performance on separate layers and check the encoded delivery file.

You can storyboard a reel with AI before anyone records a voice or animates a character. Decide what the viewer should notice when a line lands. The koosha-ai-reels-production skill asks for a still frame where the face, lettering and props each have room. Once that frame reads well, agree which elements should move and which should hold.

For an explainer, a key figure might need to stay visible while the pointer moves. For a dialogue scene, an existing pose might cover most of a line. The guide keeps new generation, client delivery and public release as separate decisions. Its included checker reads a finished file using local media tools.

What it does

  • Map a spoken line to a shot and an on-screen caption.
  • Check the edges of a speech bubble against the intended viewing size.
  • Inventory available poses before commissioning another clip.
  • Compare the picture streams after an audio-only change.
  • Distinguish technical checks from a viewing or listening review.

Install and use

  1. Begin with the approved script and artwork you can use. For delivery checks prepare Python, FFmpeg and ffprobe. You need your own editing or rendering software to make the actual video.
  2. For example, ask for three frames showing a character noticing a reminder, reading it and putting the phone down. Review the lettering before discussing motion.
  3. Export a version with audio and check its dimensions, frame count and timing against the brief. Watch it at normal speed and keep unresolved comments with that version.

FAQ

Source

Made by Koosha Studio and Hermes Agent contributors. Original skill: koosha-ai-reels-production, published on GitHub under the MIT licence.

View on GitHub